Front Burner - Minister defends Canada’s new AI strategy

Episode Date: June 5, 2026

Canada has released its long-awaited national artificial intelligence strategy. It comes as a significant portion of the country feels uneasy about what impact the technology will have. Evan Solomon, Minister of Artificial Intelligence and Digital Innovation, speaks with Jayme Poisson about AI safety and the potential for job losses.

Let's see if Toronto advisors know their life insurance providers. Hey there, who offers term plus life insurance a flexible solution with really low premiums? Oh, uh, Beneva. Correct. Who makes it easier to protect your clients with high approval rates and several built-in benefits? Veniva. You got it. Who offers whole life insurance with a whole lot of cash value? Beneva. Am I on TV?

No, not today. Looks like people are starting to know Beneva pretty well. You're stronger with the right partner, Beneva. This is a CBC podcast. Hey, everybody. I'm Jamie Pousson. Today, a feature conversation with the cabinet minister, who I think holds one of the most important jobs in the country right now. Evan Solomon, Canada's first minister of artificial intelligence and digital innovation. I have a lot of questions for him about his newly released AI strategy, so let's get straight into it.

Minister Solomon, thank you so much for coming on to Front Burner. Great to be here, James. So to start in preparation for this interview, I went back and I read some interviews with you from last summer when you were newly appointed to this job. And you sounded really quite optimistic about this technology. You talk about using it to summarize complex legislation to brief you in podcast form. You talk about being excited about a self-driving baby carriage and how you didn't want to quote, overindex on warnings and regulation. And do you still feel that way now or year on? Yeah, well, first of all, let me just say we're pretty excited to have launched this AI for

all strategy. So I just want to frame it because the AI for all is the framing device here. So I just want folks to appreciate what that means because the foundation principle from the moment I started was we got to make sure everyone's got access to something that is transformational and that it's got to be accessible. all Canadians, and it's got to serve all Canadians. And we know that this is coming. And the foundational values remain exactly the same, which is we need to make sure this technology

is used to serve Canadians for the better. And the strategy's got essentially three pillars that I've talked about all year. Trust, opportunity, and sovereignty, sovereign control. So people want to trust it. And you and I are going to talk about this. I mean, people come to me, they probably come to you and say, I'm worried about my job, my privacy, I'm worried about the future. Safety. Safety.

I mean, people worry about a lot of things. And then there's some people that are super excited about it. You know, I was, you know, with the prime minister at the launch date, you know, at Toronto General Hospital, just down the street. And they are using AI assistance in surgeries. And they're using it to make digital twins of lungs. and they made a display, every researcher that they're hiring, is using AI to enhance healthcare. So there's a lot of good.

So for sure, we need to get mass adoption. You know, we need people to use this. But people aren't going to use it, and they're not using it now. We have low levels of adoption if they don't trust it. And there's good reasons. And so the first thing we have to do is protect our kids, protect our data, and protect our privacy. And we're going to do that in this strategy.

We are going to start with that trust. And how do we protect our kids and how do we protect our data? And how do we protect our privacy is the key part of step one of this strategy, which is legislation, renewing our privacy legislation, which is forthcoming and taking action. So I want to really get into that with you because I've read the strategy and there's not a lot of details in that section. So I want to try and get some more out of you today. So as you know, in February, six children and two adults.

were shot to death in Tumblr Ridge, British Columbia. Police say Jesse Van Rootslar is responsible for a mass shooting in Tumblr Ridge. Van Rootslar died from a self-inflicted gunshot wound after killing her half-brother and mother, followed by six people at a nearby high school. In what's being described as the country's worst mass school shooting. Nothing I can say will bring your children home. The shooter's chat GPT activity was flagged by OpenAI's safety team

months before the attack and the account was banned. Federal ministers leaving a meeting with executives from Open AI say they now have more questions than answers. Was there anything troubling that you heard? Yes. What CHAP GPT said to the Tumblr Ridge shooter was not made public, but the police were not notified and there have been multiple different cases and lawsuits alleging chatbots have enabled or encouraged people to kill themselves, to kill or hurt other people. And why does this strategy not put forward specific safeguards to prevent things like this? So first and foremost on Tumblr Ridge, which remains a tragedy. People talk about it like it's in the past.

It's not. I'm a father. You know, it is not in the past, these families. And there's an ongoing legal case there. In that case to remind listeners, you know, that was a case where the alleged shooter in this case had a chat GPT account. and apparently, actually the AI flagged that there was a potential danger and the human element there that was supposed to escalate didn't. You know, I summoned immediately open AI to a meeting and I said, look, your safety thresholds and how you escalate are completely insufficient, inappropriate, and you have to take action. And we're the Canadian AI Safety Institute, which we have, which by the way, we doubled the funding for in the strategy, because,

safety is so important, and we should note that, are working on that. But so then what do you do about it? Yeah, yeah. Where are the mechanisms? Well, the mechanisms are the legislation. And Jamie, this is the key. The strategy says we're going to renew, introduce our privacy legislation, and that'll do a couple things. Treat children's information sensitively. It will give citizens, you know, this is what we're looking at. I can't give you exactly what's in the legislation, but it's forthcoming. We're looking at making sure that citizens have power against deep fakes, so they have the right to take those down. We're going to make sure that people's personal information can't be used for things like surveillance pricing. And then we are also looking

closely, my colleague Mark Miller and I, in forthcoming legislation on online harms at social media and at AI chatbots. Now, in a strategy like this, you don't list the details of the legislation, but that legislation is forthcoming, and that is step one to build the trust. So look, when I read this strategy, I did see a lot of details in here, right? Clear adoption of targets, for example, 60% of businesses using AI by 2034, 2.3 billion for training and startups, a plan to take equity stakes and high-performing AI companies. Why no details around what that legislation could look like? And I want to put some specific examples like on the table here. why not mechanisms outlined to force companies like chat GPT to be transparent with your government about what their thresholds for safety and their procedures for reporting disturbing content would be,

why not outline mechanisms to force them to do risk assessments on their products before unleashing them on the public? For example, to try to prevent the chat bot from, as was the case in this Florida school shooting last year, telling a would-be school shooter that three dead people was the unaffected. bar for widespread media attention, independent safety audits. Why isn't that coming kind of before even this AI strategy, right, so that you can build your strategy on a foundation of safety and trust? So just on the thresholds for reporting, you know, when do they escalate? You know, this is exactly what we've asked of open AI already. This is what when I summon their leadership and Sam Altman. This is why I've got the Canadian AI index.

Institute, Safety Institute, right now engaged in understanding their safety protocols, making sure they've increased their thresholds, making sure that they have contact directly with Canadian law enforcement, making sure they're going back and retroactively looking at other potential flags that they miss. But the strategy is the framework that the legislation is going to fill out. And again, we are very, this is a very detailed, as you know, a strategy on, that gives numbers, but it's one thing to give numbers that we can set in terms of how we are going to invest in small, medium-sized businesses, you know, $700 million to give small medium businesses, Canadian companies access to Canadian compute the kind of innovation they need. I can say that

in a strategy because that's a lever I control. But in legislation, in the forthcoming legislation, the details of that will be tabled in the House as they properly ought to be. They'll be debated at committee, we have been taking extensive consultation to make sure that we've got this privacy legislation that I'm going to update with these kind of things in mind and the safe social media harms that you've talked about in terms of AI chat bots and safe social media. And we are looking extremely closely to get that right. And the proper place for the details on that is in legislation and that is forthcoming. Are you certain that the Online Harms Act?

is going to tackle legislating AI chatbots? I'm not going to give you the details of what's in it until it's tabled. Do you not know the answer to that? I do know the answer to that because I'm very actively involved in what's coming. But again, it's we will table this legislation. I mean, I'm not trying to be evasive. You know, you're asking, you know, this is not to be evasive. It's that, you know, we've released an AI strategy, a national strategy.

trust is at the forefront. So, you know, I'm here to tell you, and I want to tell Canadians, that the legislation on privacy and AI safety, we're looking at all of that. You've probably spoken to Mark Miller, my colleague. We speak every day on this, and, you know, you will see the details shortly, exactly how it should come out in a legislation in the House of Commons, and there'll be a robust debate, but we are seized with making sure that our kids are safe, our privacy safe and our data safe. As the minister of AI, what would you like to see in that legislation?

Like, what would you like to see? I like how we're going to just circle from the front door to the back door, but it's a serious thing. Yeah, I'm just, I'm just, I'm sure. I'll tell you what Canadians are telling us. We did a survey on this. 11,000 Canadians gave us information of what they want to see in the national AI strategy.

the largest feedback we've ever had at the Department of ISide of industry on one topic. It was really remarkable. And they said, look, we want to make sure that when companies have our privacy, our data, we know that it's safe. Two, they would like an enforcement mechanism. So if there's a violation, there's an accountability mechanism. So they'll be enforcement. There'll be accountability. Things like deepfakes.

We hear a lot about that. hey someone put up a deep fake, I want the right to deletion, right, the right to have it taken down. So we're looking closely at that. And some folks might say, well, what have you done now? Well, we've already tabled legislation

on AI during election periods to protect our institutions and our democratic institutions. We've already tabled legislation to criminalize the non-consensual sharing of sexualized imagery, including synthetic or deep fake imagery. That should be a crucial.

criminal offense. We think that's violence against groups, especially frankly against women. And you're going to see forthcoming legislation on the privacy and the chatbots. Would you like to see a ban on chatbots for youth under 16? We're looking really closely at age restrictions, and we're looking very closely at those type of things. And as the legislation rolls out, you'll be able to see the details on it. And just, I guess this will be my final question in this section where we're we're just going around and around. But when you say it's coming, when is it coming? Because the Online Harms Act, you know, people have been waiting for it for a better part of a decade, right?

Yeah. So just so maybe folks aren't following Parliament as close as you. But so what's happened in a couple of occasions is government has put forward legislation and tabled it and then it goes to committees and it goes through the process and it actually failed. It's either died on the order paper. So twice the online harms has has been out there and twice it's failed. The privacy legislation, they've tried to do that and that's failed. So we're going again. We've recalibrated. We're looking closely.

It's a different time. Here's the difference, Jamie. People want this stuff. I mean, people want to know that their privacy are protected. People want to know that kids are protected. This is not... Yeah, it feels urgent. That's why it's funny.

Yeah, no, no, and you're right to do it. Look, look, there's two ways to answer it. Let me give you answering as a politician. You know, we're seized with it. We are. It's our obligation, our responsibility. It is. It's a absolute priority.

It is. But just the non-political answer is people want this. But you got to do it right. Like, everybody wants, you know, we've seen the harms of social media over the last 25 years. And I've raised my kids. in it. But you've got to make sure you do it right. And what do I mean by that? You got to make sure that you're not, for example, you want to give people access. You know, there's kids in rural communities or a two-s-LGB kid that is finding a community online. So you got to do it right, number one.

Number two, you don't want to cut off innovation, like in terms of the access to education, because we've got to educate young people too with this technology. So we've got to do it right. But I think we both sense and we both agree that it is an urgent thing. Yeah, and I just want to try and harness what I think, you know, many people will struggle with is that they see a strategy that's pushing adoption of AI before we've got the safety part of it, right? Well, I think the strategy is pretty clear just to articulate exactly what's in there. AI for all, the singularly first thing we're doing is trust. and the base of this is protecting privacy data and kids.

And that legislation will be the first thing that we do. I mean, this is a framework document that is doing a lot of different things. It's absolutely adoption is part of it. But we know adoption is not going to happen without the trust. So, you know, these things are happening in a sequence. Here's the plan. We should be open and transparent about the plan to people. I want people to – this is the plan.

And we're telling you this is what their first step is. And here's the next step, you know, to opportunity and to assert sovereign control. Okay. I'll just say it is the shortest section of the plan, of the strategy. Every other pillar does have quite a bit more detail. But hopefully, when the legislation comes, you will do like another two hours together. I hope that you and Minister Miller will come on together. And you can tell me even if AI will be in the Online Terms Act and then if it is how it's going to be in there.

legislation forthcoming. By the way, you know, pleasantries aside, you're right to focus on it. It is urgent. We know it. We're ready. This is why we, you know, this is, your question is not a surprise. We know the sequence here and it is forthcoming. Let's see if Toronto advisors know their group insurance providers. Oh, excuse me. Who has extensive expertise in both traditional group benefits and special risk solutions? Beniva. That's right. Who offers adaptable plans that cater to businesses big or small? Beniva. Correct. Who gives you access to the latest health trends and... I know it. Beniva. Looks like people are starting to know Beneva pretty well. I knew that too.

You're stronger with the right partner. Beniva. Have you ever wondered how clean the seats on the TTC are? I found like chicken bones or like bed bogs. Or why so many Toronto restaurant bathrooms are in dank basements? Sometimes it's the most sketchy things. Like when you go down. I was like, what is this? I'm Hayden Waters, a reporter and producer on the podcast. This is Toronto.

From breaking down Doug Ford's obsession with the island airport, we have to bring jets in. To being inside an iconic Toronto strip club in its final hours. We go beyond the headlines of the day and get to know Toronto in all its big, beautiful, frustrating, warty, fascinating glory. So find and follow us, this is Toronto, wherever you get your podcast. Let's move on to the impacts of AI on the economy. And on the labor market, this strategy includes goals to create up to

90,000 AI-related jobs for young people to support the creation of 250,000 new jobs through AI adoption. And just explain to me what these jobs are. Okay. Well, there's two different things there. We're going to create 90,000 job placements, jobs through Canada summer jobs. So we're working with our partner, EDC. And then there's organizations called My Tax, which is a well-known organization that does job placement. And what that will do is put young people in job placements. And that kind of has two purposes. Gives them jobs, gives them a salary, gives them experience.

And it also, for the businesses, helps them drive the adoption of AI because our small businesses aren't really using it and they need to find that efficiency to stay competitive. That's really important. And so we're going to create those jobs. according to the OECD, right, if you have sort of levels of adoption between 40 and 50 at the high levels of adoption, AI in the next five years will create 250,000 jobs. Now, let me just give a base so people think, where did that number come from? We already have 150,000 people in Canada working in AI jobs, not AI-related jobs in the AI industry.

Good jobs, good high-paying jobs. Overall, the digital economy is about 800,000. thousand people. So this is a job area that's growing. And so we are going to invest to make sure that we create more Canadian jobs, Canadian companies here, entrepreneurs to start companies, companies to buy Canadian technology. So we're trying to buy Canadian and ensure that the jobs stay here. But on the job question, even before you get to the 90,000 and the 250, you need to make sure that our population is educated and we have the skilled workforce. And you saw in the spring economic update, we did the 100,000 skilled trade jobs for red-sealed

workers to get young people in skill trades. The 90,000 is kind of for the digital economy as well. But we are launching a national AI literacy program. And I think this is really key. It is a big lift. We're going to make sure that every single Canadian has access to a course on AI for free because they because AI literacy is critical and then we're we'll run those through our three AI national institutes and we'll work with uh the universities and colleges and CESGEPs and um

there'll be skills training as well involved we've got to make sure we're investing in a skilled labor force and and to that end I should just say Jamie because jobs is the big question um we are setting up a uh an AI and labor and labor advisory council. So I met with most of the labor leaders. And we've been in contact with them, and we will announce soon who's on that. But to meet with me on a regular basis, to give me feedback on the jobs, because it's really important that we establish pro-worker AI. Okay. Just to that point, it did surprise me a bit to see in this strategy that there wasn't any talk of AI job destruction, just AI job creation.

basically what you just laid out and more. Why? Well, we do acknowledge very right off the top. And candidly, there's a lot of concerns about job loss. I mean, this strategy is about how do we make sure that we create conditions for job growth and that that doesn't happen. So this is real. We're not trying to sugarcoat the fact that people are concerned about job loss. I think it's in the first page. I think in my ministry or a letter, this is a 55-page document. I think that's on page one and on page two. And we have a fair bit of acknowledgement that fear of job loss is real. That's why a big part of the strategy is literacy, education.

Like what are the preventions of displacement? Skills, training, education. You got to invest in those. That's a huge part of the strategy. But look, what are the stats right now in Canada on job loss? I will just say that, I think, in the last 10 days, the Bank of Canada had a report on it. They said, so far, so far, they haven't seen any impact on job loss. That doesn't mean there won't be one.

We're not saying that that hasn't happened. We have low adoption rates. I mean, I was going to say earlier in this conversation, you talked about low adoption. I just, Daria Omoda, the head of Anthropic, has warned that AI could eliminate 50% of entry-level white-collar jobs within years. Jamie Diamond, the CEO of J.P. Morgan Chase, has said that AI will eliminate jobs and that people should stop sticking their heads in the sand. Do you think that the skills training and literacy is enough to combat a 50% elimination potentially of entry-level white-collar jobs? Well, first of all, we are building a strategy where AI is serving Canadian workers. It's a pro-worker

strategy. And that's why we've got to invest in. Now, look, I don't know what models that he's using to predict that, right? But what we're seeing is, the first and foremost, is AI is going to have impacts. There's some models, just for the record, if we want to pick models, there are some models that are saying, and there's some leaders that say, no, it's going to create lots of jobs. So, again, you know, I kind of think, there's two things that we've got to be clear-eyed about. There are two kind of camps here. There's kind of the cheerleader camp, the team pom-poms that they're like, AI is going to solve everything.

It's going to be great. Everyone's going to have the cheerleader. And then there's the pitch for a crowd. Stop building data centers. We should stop. We are not that. Like, we are going to be pragmatic. That means educating a workforce, getting literacy, investing in Canadian companies, investing in Canadian

innovation, keeping it here, building data centers that the right way, sustainably. with Canadian values. And we've got to have some control over this. I mean, we are not passive here. We have to build here in Canada, regulate here in Canada for our kids, and we are not passive agents. We have a sense of control over this,

and that's good. Is being pragmatic, does that also not require you to plan for all scenarios? So Nick Frost, the co-founder of Cohere, an AI company that the Canadian government champions all the time says AI will make inequality worse. He also said that we need to build other social nets for employment. And I just, I don't see anything about improvement to social

programs like unemployment insurance in this document. I don't see anything about the development of universal basic income. I actually don't see anything that the government is taking seriously the idea that there could be serious disruption to the labor market. So, so, I, I, Let me push back because we just literally had a conversation about launching a national education program, investing in 90,000 jobs for young people. I'm, and I just don't want to dismiss it, but I literally see nothing. But that is the safety net. Yeah. Well, I mean, education is important.

I mean, I'm just, I think it's really important. I mean, I have two kids, you know, the best insurance that you give your kids, is obviously a safe, healthy environment. That's the cornerstone of our strategy, safe trust. And how do you give them opportunity? You make sure they've got access to the tools they need, education, and access to the tools. And we're going to give young people access,

every post-secondary student access to the tool. So we've got to make sure that there's not this discrepancy. So on one hand, you know, you don't have the tech haves and the tech have-nots. And then we've got to give them the opportunity to build, and to create. And we're investing in workers in this strategy. I mean, this idea, you're asking me about a scenario that you've posited. I mean, not me.

You know, I know, I know. Some of the smartest, the smartest people. I get it. But our foundational goal, our financial belief is we ought to invest in Canadian workers. We've got to invest in Canadian young people, which we're doing. And you see it. There's, you know, there are billions of dollars in this strategy that are investing in Canadian companies, small, medium, size, 95%, 98% of our companies are small, medium-sized companies.

We're investing in them. They got to get access to capital so they can grow and the jobs are here. And just so, I just want people to get this. I mean, we're gliding over a lot of big topics here. Jamie and I know there's a lot in here, but you said cohere. So just like, I'm just going to double click on that for a second. just so folks get it. And maybe they do.

But, you know, there are only four countries in the world that have these large language models that we're talking about. Now, you mentioned, you know, anthropic and open AI. So, you know, the United States has a bunch and China's got a bunch. France has one. The only other country in the world that has one is Canada. It is cohere. It's just an ideal to partner with a German, a smaller German company so they can grow.

But that gives us strategic options. So Canadians can have control and Canadian control. And, you know, we need agency to build this in a responsible, reliable way. And, you know, I would just say that, you know, most of our conversation, which is great, because this is part of what the strategy is, is about what AI is. What could it be? But we should also acknowledge what AI does. I mean, we're literally at a hospital this morning. I spent three hours there, and I met dozens of doctors who are doing surgery on cancer surgery.

I literally met a team who are using AI to do cancer surgery to significantly less invasive, to find the cancer to reduce second surgeries. I mean, we should also acknowledge that this is going to serve at its best. That doesn't mean that's all we should focus on, but we've got to be open to the opportunities that this is creating. and candid about the concerns, which I think this strategy, that's what I mean when I say pragmatic. Yeah, I just, I think maybe it's worth flagging. Like, I do totally see how AI is going to be really beneficial in areas like cancer research and in what you just laid out. I'm not trying to say that it's not useful in lots of ways. I also do want to flag that Ontario's Auditor General just put out a report about AI note-taking tools intended for Ontario doctors that were hallucinating and correct information.

You know, hallucinations are like a persistent issue with these tools and there's all these problems. And I just kind of, to end this conversation that we're having to say, I wish we had more time, but thank you for being so generous. You know, you have talked about how you think this technology is powerful and transformative and revolutionary. I've seen you before compare it to the invention of the printing press. And I just, I would posit that if that's true, then we, you and I, you and I, you and I, I don't know what it is fully capable of. And do you worry that we're racing so quickly towards the widespread use of this thing that we don't even truly know the consequences of?

So I have compared it to like a Gutenberg moment. Yeah. Because it's a transformational technology. You know, the Internet was a transformational technology. And we lived through that. And that was my background, building a company around that. And here we are with another transformative technology. And the first point is you're right.

We, you know, this is our framework for AI for all is a principled approach to how we are approaching this transformative technology. We think first and foremost, you know, you've got to establish your principles and then your action items. And our principles are start with trust and build trust, create opportunity and make sure that this is good for our sovereignty and our sovereign control. But there is an element of humility here. I mean, this strategy is built to adjust because there's going to be new opportunities and there's going to be new challenges. And there's challenges about security. I mean, we've talked about, you know, mythos, this powerful technology that, you know, I'm glad to say Canada now has access to through our Canadian Cybersecurity Center, which is really important. But it's a, you know, Canadians are rightly concerned.

Are these things so powerful that they will break through our security systems? So this is why we're working with our safety institute and our cybersecurity center to handle that. But we got to approach it with humility. No one country can control everything. But it's not an on-off switch. It's not well. If you can't control it, just give up or stop. It's assert control.

Protect your kids. Make sure that there's, you can, as you talked about, make sure we can at least be available for the benefits of this. And make sure that, you know, we have a sense of control over how this rules out. And this strategy is part of that. And as new challenges assert themselves, we will keep working. But this is the start of a very important investment in giving us the tools to build the kind of country with these tools that we want. Okay.

Minister Solomon, thank you very much for being here. Thanks, Jamie. All right, that is all for today. Front burner was produced this week by Matthew Almaha, Joytha Shen Gupta, Mackenzie Cameron, Kevin Sexton, Kieran Outshorn, Kristen DeJager, and Mitchell Stewart. Our YouTube producer is John Lee. Our music is by Joseph Shabbison. Our senior producers are Imogen Burchard and Elaine Chow.

Our executive producer is Nick McCabe Locos, and I'm Jamie Poisson. Thanks so much for listening. Talk to you next week. For more CBC podcasts, go to cbc.ca.ca slash podcasts.